Southern Japan and the Carribean are both endemic for human T-lymphotropic virus type I (HTLV-I) infection. In these areas, however, epidemiology of HTLV-I appears to differ. Various observations from population-based studies indicate that differences in host immune response to virus infection is likely to result in different disease manifestations across geographic regions. While gender, age and route of infection may partly determine host immune response, the observed geographic differences may also reflect, in part, host genetic background as evidenced by the distribution of human leukocyte antigens (HLA) and other polymorphic genes in addition to the presence of environmental factors. Viral Epidemiology Branch has been conducting population-based studies in the Caribbean to understand risk factors for HTLV-I infection and its disease pathogenesis. [unreadable] [unreadable] In 1985-6, the Food handlers Study, a cross-sectional survey of island-wide HTLV-I prevalence in Jamaica was conducted to determine the seroprevalence of this virus. Over 13,000 applicants for food handling licensure were screened for HTLV-I. Overall HTLV-I seroprevalence in Jamaica was 6%, with the prevalence higher among women and those who are older. Approximatley 400 HTLV-I-infected and -uninfected persons from this population were enrolled in a case-control study in 1987-8, which investigated risk factors for HTLV-I infection. Some of these participants returned for the third follow-up in 1992. This study is a basis of our subsequent investigations of polymorphic genes and susceptibility to HTLV-I infection and virus markers. [unreadable] [unreadable] Jamaica Mother-Infant Study is a prospective cohort study enrolling 350 women and their children since 1989. Twenty-eight children became HTLV-I positive during the first two years of follow-up. Many children, both HTLV-I-positive and -negative, have been followed over 10 year period with questionnaire on interval health history, physical examination and phlebotomy. The study identified maternal low socioeconomic status, longer breast-feeding, rupture of membrane during labor, as the risk factor for HTLV-I transmission. High maternal provirus loads and antibody titers were independent risk factors for infection in children. HTLV-I infection in children was associated with a higher incidence of dermatitis, severe anemia, lymphadenopathy and presence of abnormal in the peripheral blood. Children with these symptoms had a higher HTLV-I proviral load as compared to those without symptoms. [unreadable] [unreadable] We are also following a cohort of 66 blood transfusion recipients since 1987, who have been exposed to HTLV-I-positive blood product before the HTLV-I screening was in place, for disease incidence and mortality. We observed 24 seroconversions in the recipients, with a sharp initial elevation and subsequent stabilization of HTLV-I titer and provirus load. One seroconverters developed HAM/TSP, a chronic progressive myelopathy associated with HTLV-I. There seems to be a 3-fold excess risk of skin rash among seroconverters as compared to non-seroconverters. [unreadable] [unreadable] Also in mid 1980s, University of West Indies, Jamaica and two referral general hospitals in Trinidad & Tobago established parallel registries of hematologic malignancies. A hospital-based case-control study of non-Hodgkin's lymphoma (NHL) was then conducted using cases identified through these registries. Persons with NHL, particularly T-cell lymphoma, were significantly more likely to be HTLV-I positive. Up to 60-70% of NHL cases in the Caribbean were estimated to be attributable to HTLV-I. Further characterization of adult T-cell leukemia (ATL) was also performed using case data from these registries. [unreadable]